Variation of atmospheric stability in a typical subtropical forest and its effects on potential evapotranspiration
Atmospheric turbulence structure and its scalar transfer capacity,generally described using the concept of aerodynamic resistance(denoted as ra)in evapotranspiration models,is one of the key factors affect-ing the accuracy in estimating potential evapotranspiration(denoted as PET).The aerodynamic resistance is significantly affected by the atmospheric stability,however,due to that the atmospheric stability data are not usually available,most of the previous studies estimated ra based on the neutral hypothesis,which would result in uncertainties.This study investigated the atmospheric stability and evaluated the errors brought by the neut-ral hypothesis in PET estimation in a subtropical forest in southwestern China,based on EC measurements and Monin-Obukhov similarity theory.Results showed that,the neutral,stable and unstable atmospheric stratifica-tions exhibited explicitly diurnal and seasonal patterns,with unstable stratifications being predominant at noon,stable stratifications being more common in the morning and evening,and neutral stratifications being distrib-uted throughout the daytime.Moreover,the high-frequency periods of unstable stability are longer in spring and summer but shorter in autumn and winter.Overall,the mean ratio of unstable,neutral and stable stratifica-tions during the experimental period was 41.4％,34.7％and 24.0％,respectively.Additionally,the aerodynam-ic resistance estimation based on the neutral assumption tends to overestimate ra,while neglecting the impacts of atmospheric stability leads to underestimation of PET with significant errors(RMSE=25.82 W/m2),and the deviation of the error becomes more pronounced with increasing PET.Finally,potential evapotranspiration based on the relationship(denoted as PET equ)between the equivalent aerodynamic resistance(denoted as ra)and wind speed(R2=0.50,P＜0.05,N=394)are more accurate in simulation(R2=0.95,RMSE=20.70 W/m2),compared with that based on neutral hypothesis(R2=0.92,RMSE=25.82 W/m2).Important to note that the rela-tionship between(r)a and wind speed would differ dramatically in different regions,further studies are still needed.These results can provide scientific basis for ecological hydrological modelling,water resource man-agement,and drought assessment.
NETL
NSTL
万方数据
